Evaluating State and Local Business Incentives Cailin Slattery, - - PowerPoint PPT Presentation

Evaluating State and Local Business Incentives

Cailin Slattery, Columbia GSB Owen Zidar, Princeton & NBER

Brookings Municipal Finance Conference

July 14, 2020

Motivation

• State and local governments spend billions of dollars each year on tax incentives

and subsidies to attract and retain firms (Bartik, 2017, Slattery, 2019)

• Incentive policies are highly controversial
• Attracting industrial activity is key for local economic growth and prosperity
• Others question incentive spending effectiveness and mounting costs
• Evaluating these incentives requires overcoming three challenges
• 1. Data limitations: difficult to measure prevalence, size, and composition of incentives
• 2. Lack of transparency: hard to determine selection process
• 3. Do not observe how economic activity would have evolved in the absence of deals
• New data on incentives enable us to make progress

1 / 18

What we do in this paper

• 1. Characterize these incentive policies
• 2. Describe the selection process that determines which places give incentives and

which firms receive incentives

• 3. Evaluate the economic consequences and discuss policy implications

2 / 18

• 1. The State and Local Business Incentives Landscape

Three Business Tax Instruments

• 1. Lowering the corporate tax rate: Lowers tax bill for all C-corps, encouraging

entry of new firms and expansion of existing firms

• 2. Narrowing the corporate tax base: Lower tax bill for set of firms, based on

activity/industry. Encourages entry of new firms in that industry/increase in targeted activity

• 3. Offering firm-specific tax incentives: Offer one firm a subsidy for their

commitment to locate in the jurisdiction and create a certain level of employment and investment

3 / 18

States take different approaches

Average AL CA NV NY PA SC TN WV Instrument 1: Corporate Tax Rate (%) 6.5 6.5 8.8 7.1 10 5 6.5 6.5 Corporate Tax Revenue Per Capita ($) 162 90 246 264 193 81 193 118 Instrument 2: Tax Credits per capita ($) 19 11 60 33 15 32 16 Econ Development per capita ($) 34 15 2 5 142 25 8 35 177 Instrument 3: Number of subsidies 14 15 13 4 20 3 16 12 4 Cost per job ($) 45,785 12,466 4,997 42,339 11,712 93,406 6,433 11,805 34,345 Incentives as a percent of Corp Tax Revenues (%) 38 29 25 N/A 66 20 49 26 150

4 / 18

States take different approaches

Average AL CA NV NY PA SC TN WV Instrument 1: Corporate Tax Rate (%) 6.5 6.5 8.8 7.1 10 5 6.5 6.5 Corporate Tax Revenue Per Capita ($) 162 90 246 264 193 81 193 118 Instrument 2: Tax Credits per capita ($) 19 11 60 33 15 32 16 Econ Development per capita ($) 34 15 2 5 142 25 8 35 177 Instrument 3: Number of subsidies 14 15 13 4 20 3 16 12 4 Cost per job ($) 45,785 12,466 4,997 42,339 11,712 93,406 6,433 11,805 34,345 Incentives as a percent of Corp Tax Revenues (%) 38 29 25 N/A 66 20 49 26 150

Today, I will focus on firm-specific tax incentives.

4 / 18

Why offer a firm-specific tax incentive?

• 1. Attract a “high-benefit” firm
• 2. Can contract with firms on investment and hiring
• 3. Don’t have to lower revenue collected from all firms in the state
• 4. Target mobile firms, raising revenues more efficiently (Ramsey 1927)
• 5. Retain a valuable firm

5 / 18

Why offer a firm-specific tax incentive?

• 1. Attract a “high-benefit” firm
• 2. Can contract with firms on investment and hiring
• 3. Don’t have to lower revenue collected from all firms in the state
• 4. Target mobile firms, raising revenues more efficiently (Ramsey 1927)
• 5. Retain a valuable firm

Costs to using firm-specific tax incentives:

• 1. It is hard to pick winners
• 2. It is hard to know if firm is inframarginal
• 3. Lack of transparency leaves incentives exposed to political capture
• 4. Most distressed places may not be able to afford to compete
• 5. Providing generous incentives requires raising revenue from other taxes

5 / 18

What do we mean by firm-specific incentive?

2008 Volkswagen Deal in Tennessee

• VW chooses Chattanooga for new assembly plant
• Promises 2,000 emp and $1B investment
• TN grants VW a subsidy worth $558 million
• Local property tax abatements over 30 years ($200M)
• Enhanced state job and investment tax credits over 20 years ($200M)
• Property given to VW ($81M)
• Worker training ($30M)
• Highway and road construction ($43M) + Rail line upgrades ($3.5M)
• TN promises specialized tax credits for any neighboring suppliers
• TN projected VW would have $100M in annual payroll, help create 14,000 total

jobs, and have a total economic benefit of $600M per year

6 / 18

Volkswagen location decision was “truly a very close competition”

• Initially considered “more than 100 candidate sites”
• Runner-up in Huntsville, AL, subsidy offer at least $386 million

7 / 18

Volkswagen location decision was “truly a very close competition”

• Initially considered “more than 100 candidate sites”
• Runner-up in Huntsville, AL, subsidy offer at least $386 million

Employment in Auto Manufacturing Diff Between Winner and Runner-up

2000 4000 6000 8000 10000 Employment in NAICS 336

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Hamilton, TN Huntsville, AL

• 10000
• 8000
• 6000
• 4000
• 2000

Employment in NAICS 336

2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

3854 workers

7 / 18

• 2. Data on State and Local Business Incentives

Firm-Specific Incentives: Data Collection and Sources

Company Year Winner Runner-up Subsidy ($M) Jobs at Stake Invest ($M) Hyundai 2002 Montgomery, AL Hardin, KY 234.6 2,000 1,000 Fidelity Investments 2006 Wake, NC Duval, FL 88.2 2,000 100 American Greetings 2011 Cuyahoga, OH Cook, IL 117.2 1,700 25 Procter & Gamble 2015 Berkeley, WV Franklin, PA 21.0 700 500 Sources: Subsidy Tracker + Site Selection

Example

Articles on Subsidy Deals

Example

Tax Expenditure Reports

Example

State Budget Documents

Example 8 / 18

Firm-Specific Incentives: Data Collection and Sources

Company Year Winner Runner-up Subsidy ($M) Jobs at Stake Invest ($M) Hyundai 2002 Montgomery, AL Hardin, KY 234.6 2,000 1,000 Fidelity Investments 2006 Wake, NC Duval, FL 88.2 2,000 100 American Greetings 2011 Cuyahoga, OH Cook, IL 117.2 1,700 25 Procter & Gamble 2015 Berkeley, WV Franklin, PA 21.0 700 500 Sources: Subsidy Tracker + Site Selection

Example

Articles on Subsidy Deals

Example

Tax Expenditure Reports

Example

State Budget Documents

Example

543 subsidies, average $178M, 1,500 jobs (2002-2017)

8 / 18

Most subsidies go to manufacturing, technology, and high-skilled services

# of Subsidy Jobs Cost Per Investment Deals ($ M) Promised Job ($) ($ M) Full sample 543 178.4 1,487 119,972 757.5 Automobile manuf. (3361) 56 293.6 2,768 106,057 854.8 Aerospace manuf. (3364) 31 585.8 2,734 214,237 534.5 Financial activities (5239) 25 92.3 2,652 34,809 286.8 Scientific R&D svc (5417) 22 113.7 518 219,259 185.0 Basic chemical manuf. (3251) 18 187.4 196 956,701 779.0

Notes: This table includes the mean deal characteristics (subsidy size, jobs, investment) for select industries. Dollars in 2017 USD.

• Automobile manufacturing most popular, has largest expected multiplier
• Policymakers target firms with large agglomeration effects (high-benefit)
• 10 industries make up 47% of sample

9 / 18

Subsidies go to large establishments, from large profitable firms

Small establishments are less likely to receive discretionary subsidies:

All Discretionary Subsidies (2002-2017) Jobs promised # Subsidies Estab Entry % Coverage 1 - 99 39 8,971,339 0.00 100 - 249 47 26,126 0.18 250 - 499 80 4,251 1.88 500 - 999 141 1,419 9.94 1000+ 236 639 36.93

10 / 18

Subsidies go to large establishments, from large profitable firms

Small establishments are less likely to receive discretionary subsidies:

All Discretionary Subsidies (2002-2017) Jobs promised # Subsidies Estab Entry % Coverage 1 - 99 39 8,971,339 0.00 100 - 249 47 26,126 0.18 250 - 499 80 4,251 1.88 500 - 999 141 1,419 9.94 1000+ 236 639 36.93

Compared to publicly traded companies, subsidized firms are larger, more profitable:

All Compustat Subsidized Firms Subsidized Firms: Year of Deal Mean Median Mean Median Mean Median Employees (1000s) 9.0 0.6 72.0 34.2 100.9 64.3 Capital Stock ($M) 1,514.4 28.2 12,098.3 3,004.6 18,865.2 7,720.1 Gross Profit ($M) 1,139.8 67.5 13,239.3 4,007.9 20,743.3 8,969.8 Market Value ($M) 2,997.1 189.5 45,988.1 13,305.6 76,582.2 27,924.0 Observations 107,219 2,470 313 10 / 18

Low-wage locations provide more generous subsidies

Albany, NY (2008) Suffolk, MA (2016) San Francisco, CA (2017) Charleston, SC (2004) 200 400 600 800 1000 Cost per job (1000s 2017 USD, binned) 40 60 80 100 120 Average Wages (1000s USD, binned)

Notes: Average wages are measured in the year of the firm-specific deal. Triangles in plot are individual data points; circles are binned data. Best fit line estimates are taken from population-weighted linear regression of cost per job on average wages. Winning counties are richer than average 11 / 18

• 3. Effects of State and Local Business Incentives

Event Study: Within-Industry Employment Effects of Subsidy

• 1000

1000 2000 3000 County-level employment

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal Impact 95% CI

Notes: This figure shows the event study estimates of the effect of winning a firm-specific deal on county level employment within the NAICS 3-digit industry of deal

12 / 18

Event Study: Spillover Effects of Winning a Subsidy Deal

• 4000
• 2000

2000 4000 Employment

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal 3-D 2-Digit Residual 1-Digit Residual

Notes: This figure shows event study estimates of the effect of winning a firm-specific deal on three outcomes: employment in 3-digit industry of deal, 2-digit residual employment, and 1-digit residual employment.

13 / 18

Comparison with Prior Results

We detect direct effect of winning a firm on local employment within industry of deal

• Don’t find effects on employment outside the directly affected industry

Previous studies have found more support for spillover effects using estab.-level data

• “Million Dollar Plants” data: 82 subsidy deals from Site Selection Magazine,

mostly manufacturing, in 1980s and 90s (Greenstone & Moretti 2003)

• Greenstone, Hornbeck & Moretti (2010) find substantial TFP spillovers for estabs

in winning areas (Patrick (2016) finds more modest spillover effects)

• Our data selected on subsidy size, MDP data selected on size of investment

MDP Results

14 / 18

• 4. Policy Discussion

Improvements to current policy

Policymakers can design incentives with equity considerations in mind

• Target tax incentives to hard-hit regions (Bartik 2019)
• Subsidize employers who promise to hire local residents (Bartik 2019)
• Target marginal investments and job creation for high-multiplier industries
• More transparent and less political than incentives for individual firms
• Should also evaluate whether incentives “trickle down”

Need more rigorous evaluation and transparency requirements

• Federal accounting rules require incentives reporting, but reports are low quality
• Few states require systematic evaluation of incentive programs

15 / 18

Considering harmonization

Various U.S. governors have proposed “truce” on subsidy competition

• NY lawmakers propose “End Corporate Welfare Act”, urge others to sign on
• This is more attractive to NY than in more distressed regions

Harmonization may be more effective at the regional level (e.g. Kansas City)

• However, both states rushed to finalized deals right before truce enacted

16 / 18

A larger (super-)federal role

Centralized approach may be key to avoid competition and address equity concerns

• Federal programs with lenient requirements unlikely to be effective
• E.g. Opportunity Zone program in the U.S.
• Empowerment Zones increased employment in places with poverty rates above

40%, unemployment over 15% (Busso, Gregory and Kline 2013)

• Recent Proposal: Tennessee Valley Authority for the 21st century

(Austin, Glaeser and Summers 2018)

The EU restricts state aid to reduce concerns about tax competition

• Structural funds implemented at super-federal level
• Encourage investment, capital deepening, and econ development in distressed

areas, to reduce regional disparities

17 / 18

Open questions and directions for future research

• 1. How much do state and local business tax incentive policies improve the

well-being of underemployed and low-income workers?

• As argued in Amazon HQ2 NYC case, will all good jobs go to migrants and

high-skilled, leaving locals with more congestion and higher prices?

• 2. How effective are these approaches relative to other policies?
• Places could instead invest in education, amenities
• 3. How big of a role does politics play in subsidy-giving?
• Distribution of resources within a state?
• Actual effects on re-election success?
• 4. Does targeting subsidies at the largest firms have anti-competitive effects in the

product market?

• Implications for conversation on increasing industrial concentration?

18 / 18

Thank you!

States with high corporate tax rates have narrower bases

IN VA NY NC CA 2 4 6 8 10 Corporate tax rate (%) 20 40 60 80 Per capita tax expenditures ($)

1 / 9

Back 2 / 9

3 / 9

Back 4 / 9

Back 5 / 9

Back 6 / 9

Which Places Provide Firm-Specific Subsidies?

Unique counties in 2000

County: Winner (Full) Winner (Analysis) Runner-up Average Pop > 100K Mean Median Mean Median Mean Median Mean Median Mean Median Employment (K) 201.3 82.0 229.1 142.9 303.5 157.5 44.1 11.3 197.1 102.9 Population (K) 407.0 171.2 453.9 285.1 610.0 308.1 90.9 25.2 400.8 208.6 Average wages (K) 45.5 42.8 48.6 45.0 48.6 45.0 34.8 33.1 44.4 42.1 Population density 1,096.7 285.3 1,524.9 485.2 1,702.1 506.3 229.4 42.4 1,088.5 341.0 % emp in mfg. 21.5 18.0 20.1 16.0 17.6 15.3 19.3 17.3 16.4 14.8 % emp info & prof svcs. 19.3 17.2 22.4 22.1 24.1 24.0 9.6 8.3 21.2 19.6 % urban 73.2 78.5 81.0 90.6 82.8 91.8 39.1 38.4 80.4 85.0 % Bachelor’s or more 22.1 20.3 25.4 24.6 26.9 25.4 16.5 14.5 24.9 23.3 % white 78.1 81.1 77.4 79.2 75.7 77.8 84.5 91.3 79.5 83.0 % Hispanic 7.0 3.1 8.1 3.9 8.1 3.9 6.2 1.8 9.1 4.3 % foreign-born 6.2 3.5 7.7 4.7 8.5 5.5 3.5 1.7 7.7 5.2 log housing units 11.2 11.2 11.5 11.6 11.7 11.7 9.4 9.3 11.5 11.4 Wage bill (M) 10,969.5 3,403.9 12,789.2 6,751.4 17,477.6 7,689.0 2,086.8 376.7 10,059.3 4,207.9 Personal income (M) 19,640.2 6,592.2 23,161.7 11,790.5 31,131.8 14,512.0 3,968.0 792.9 18,809.3 8,473.0 Personal inc/capita (K) 40.9 39.3 44.5 41.8 45.6 41.7 34.1 32.8 42.9 40.6 Unemployment rate (%) 4.0 3.7 3.7 3.4 3.8 3.5 4.4 4.1 3.9 3.6 Observations 268 115 126 3,107 537

Back 7 / 9

Comparison: Within-Industry Employment Effects of Subsidy Deal

Our Subsidy Deals Dataset Million-Dollar Plants Dataset

• 1000

1000 2000 3000 Employment in 3-D industry of deal

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal Impact 95% CI

• 2000
• 1000

1000 2000 Employment in 3-D industry of deal

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal Impact 95% CI

Notes: This figure plots event study estimates of the effect of winning a firm-specific deal on employment in NAICS 3-digit industry of deal.

Back 8 / 9

Comparison: Spillover Effects of Winning a Subsidy Deal

Our Subsidy Deals Dataset Million-Dollar Plants Dataset

• 4000
• 2000

2000 4000 Employment

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal 3-D 2-Digit Residual 1-Digit Residual

• 5000

5000 10000 Employment

• 5
• 4
• 3
• 2
• 1

1 2 3 4 5 Years since deal 3-D 2-Digit Residual 1-Digit Residual

Notes: This figure plots event study estimates of the effect of winning a firm-specific deal on on three outcomes: employment in 3-digit industry of deal, 2-digit residual employment, and 1-digit residual employment.

Back 9 / 9